The broad objective of the Resource is to apply advanced methods in Computer Science, particularly in Artificial Intelligence (AI), to biomedical problems. The Resource will promote the development and use of computer systems for intelligent consultation in medical diagnosis and therapy and for research assistance in processes of scientific experimentation and theory formation. The organization of knowledge in a domain, both formal and judgmental (obtained from experts), and its representation in computers, is of central importance in the Resource. Knowledge-based systems that rely on models, heuristic rules, and other representations that facilitate the use of knowledge in various reasoning tasks, will be developed and studied. The Resource will have three major areas of study: Area 1 - Medical Modeling and Decision Making in ophthalmology, with emphasis on glaucoma, visual optics and visual field loss; Area 2 - Modeling of Belief Systems and Commonsense Reasoning with emphasis on the psychology of action interpretation; and Area 3 - Artificial Intelligence studies, with emphasis on Representations, Reasoning and System Development problems of relevance to the Resource objectives. The Resource will continue to sponsor the national Artificial Intelligence in Medicine (AIM) workshops, and it will increase its efforts in AIM dissemination and training activities. The RUTGERS-10 computer will be enhanced, and it will serve as a shared resource, in coordination with the SUMEX-AIM facility, for the national AIM community. Our research is interdisciplinary and interinstitutional. The studies in Medical Modeling and Decision Making will be performed jointly by computer and medical scientists at Rutgers and Mt. Sinai, in collaboration with a national Ophthalmological Network (ONET) of investigators in seven medical institutions. The AIM computer network will support these collaborations. Methods will be developed for translating prototypes that run on minicomputers linked to the large AIM computers.